Artificial intelligence (AI), machine learning, and generative AI are increasingly driving innovations across various sectors. The field of patent research significantly benefits from these technologies. INTERGATOR Patent Search enables semantic searches that identify similar documents based on text inputs, eliminating the need for users to formulate complex Boolean queries.
Hybrid Search Strategy for More Precise Results
However, a purely AI-based search that relies on vector models can encounter limitations, often yielding numerous irrelevant results. Therefore, INTERGATOR employs a hybrid strategy: combining AI-based semantic search with Boolean filters. Boolean filters allow for targeted narrowing of results to exclude irrelevant hits, while semantic search facilitates the identification of similar and potentially relevant documents despite varying terminology. This iterative optimization process — where identified relevant hits refine the semantic query and adjust filters — already delivers excellent outcomes. Users are supported by user-friendly assistants that enable transparent query formulation and quick adjustments to the search strategy.
Generative AI: The Next Step
The Future of Patent Research
With the upcoming release of INTERGATOR Patent Search in the first quarter of 2025, optional generative technologies will become available. This provides companies with a powerful solution to further enhance their innovative capabilities. Generative AI elevates patent research to a new level, aiding in answering complex questions more simply and efficiently.
A standout feature of the large language models (LLMs) underpinning generative AI is their application in analyzing and transforming texts. For instance, technical patent documents can be restructured to present information in clearer language that is more accessible to engineers. This approach departs from the traditional style of patent documents, facilitating development engineers’ evaluation of research results.
Additionally, generative AI enables graphical representations of functional and component analyses. These visualizations provide an intuitive means to comprehend interactions between various components and to grasp technical relationships more swiftly. This not only enhances research efficiency but also simplifies the comparison of publications.
Another application is the specific assessment of search hits. By comparing an original document with found documents, the relevance of individual hits can be better evaluated. This function assists in more precisely assessing innovations, analyzing the patentability of an idea, or identifying grounds for opposition.
Further Applications of Generative AI
INTERGATOR Patent Search expands AI capabilities through the integration of generative technologies. These not only allow for deeper analysis of publications and search results but also offer numerous innovative application scenarios:
- Improving Search Result Relevance: Targeted analysis allows for quicker identification of pertinent documents.
- Translation (EN ↔ DE): Accurate translations facilitate working with international patent documents.
- Creation of Research Reports: Automated reports summarize findings clearly and concisely.
- Image-Based Search (Image → Query): Expands research options significantly with image-driven queries.
- Publication Analysis (Result Evaluation): Accelerates the identification of relevant publications through qualitative assessments.
- Recognizing Ideas for Further Innovation: Leverage search results to develop new approaches.
- Chatbot for Research & Development: Provides direct support for research and answering queries.
- Patentability Research (in a Secure Environment): Identifies critical information for evaluating patentability.
- Transforming Patent Language into Developer Language: Simplifies complex texts for a technical audience.
- Comparing Publications: Enhances the efficiency of analyzing similarities and differences.
- Research in Original Documents (Without Translation): Enables direct work with texts in their original language.
- Drafting Patent Applications: Creates initial drafts for patent filings.
Security and Data Protection as Top Priorities
A key feature of INTERGATOR Patent Search is the deployment of generative AI on company-owned servers within a German data center. This setup guarantees the highest level of data security and ensures that user inputs are protected and cannot be accessed or processed by third parties. This strict data sovereignty is essential, particularly in sensitive areas such as patent research, as it builds trust and helps companies safeguard their innovations and trade secrets.
